Magnetoresistive (MR) read transducers are employed extensively for reading magnetically recorded data, typically recorded in parallel tracks on magnetic disk or on magnetic tape. A key dimension for MR transducers is the width of the active sense region, which defines the trackwidth of the recorded data that is read. If the active sense region width is narrower and more precisely defined, the data tracks of recorded data may be narrower and closer together, thereby allowing more tracks on the same dimensioned recording media and thereby increasing the data capacity of the recording media.
Anistropic magnetoresistive (AMR) heads are one type of magnetoresistive head and include an AMR sensor contacted by longitudinal bias and conductor leads films. Typically, the AMR sensor and bias and conductor films are sandwiched between inductive write films, or between an inductive write film and a shield.
The AMR sensor typically includes a ferromagnetic sense layer separated by a nonmagnetic spacer layer from a ferromagnetic reference layer. The reference layer generates a transverse bias reference field, either by activation from the current supplied to the sensor for sensing the resistance of the sense layer, or by being pinned in a transverse direction through exchange coupling between the reference layer and an underlying ferromagnetic pinning layer. The transverse bias is conventionally used to maintain the sensor in its most linear operating range.
Longitudinal bias is conventionally required to reduce Barkhausen noise for the stabilization of MR sensors. Two major types of longitudinal bias are currently utilized, the overlaid longitudinal bias wherein separate longitudinal bias layers overlay the sense layer in each of the end regions, and the abutting longitudinal bias wherein the sense layer is only in the center region and the bias layers abut and are adjacent the sense layer in the end regions.
Another magnetoresistive head gaining in usage is the giant magnetoresistive sensor (GMR) which employs pinned and free ferromagnetic layers separated by a thin film layer of nonmagnetic material. The voltage across the GMR sensor is related to the rotation of the magnetization in the free ferromagnetic layer as a function of the magnetic field being sensed. Although the mechanism for sensing the magnetic field of the data being sensed is different from the magnetoresistive effect of the AMR sensor, the GMR sensor must also be stabilized by a longitudinal bias field. Typically, the GMR sensor is stabilized by overlaid longitudinal bias layers at each end of the sensor, although abutting longitudinal bias is also employed.
The longitudinal bias layers are conventionally overlaid with conductors which comprise the conductor leads at each end of the sensor. A current is supplied between the conductors which is conducted by the longitudinal bias layers to each end of the sense layer(s) and the voltage generated by the current across a central region of the sense layer(s) between the bias layers is the sense signal representative of the sensed recorded data. Thus, the active width of sense layer is defined by the central region between the longitudinal bias layers at each end of the sense layer central region.
Theoretical drawings of MR heads indicate that the edges of the longitudinal bias layers and the leads are perfectly aligned and perfectly vertical, either in the overlaid longitudinal bias or the abutting longitudinal bias. Conventional manufacturing produces edges that are less than perfectly aligned and less than perfectly vertical. In a mass production environment, the design of the disk file must accommodate the various active areas of MR read transducers. Therefore, the more difficult it is to align the edges, the more difficult it is to define the width of the active sense region, and the parallel tracks of recorded data must be located further apart to accommodate the various widths of the active sense region of the manufactured transducers.
FIG. 1 illustrates a conventional overlaid MR transducer, which may either be an anisotropic MR transducer or a giant MR transducer, with overlaid longitudinal bias layers in the end regions. A plurality of sensor layers 10 are provided. In an anisotropic MR transducer, the sensor layers 10 may include a pinning layer and a soft adjacent layer (SAL) which together provide the transverse bias, a spacer layer, and the sense layer. In a giant MR transducer, the sensor layers 10 may include a pinning layer and a ferromagnetic layer fixed by the pinning layer, a nonmagnetic spacer, and the free ferromagnetic layer. A separate overcoat layer 11 is provided to protect the sensor 10.
Longitudinal bias layers 12 and 13 and conductor lead layers 14 and 15 overlie the sensor layers 10 in the end regions. Typically, the longitudinal bias layers are exchange-coupled ferromagnetic/antiferromagnetic films. As can be seen from the illustration, the longitudinal bias layers 12 and 13 and the conductor lead layers 14 and 15 of the conventional overlaid MR transducer are at a low angle and do not provide easily aligned and vertical edges to define an active central region of the sensor layers 10.
FIG. 2 illustrates a conventional abutted MR transducer, which may either be an anisotropic MR transducer or a giant MR transducer, with the sensor layers 20 abutted by longitudinal bias layers 21 and 22 in the end regions. Conductor lead layers 23 and 24 overlie the longitudinal bias layers 21 and 22 in the end regions. Again, the conductor lead layers 23 and 24 of the conventional abutted MR transducer are at a low angle and do not provide easily aligned and vertical edges to define an active central region of the sensor layers 20.
The reason that the conventional MR transducers do not provide easily aligned and vertical edges to the longitudinal bias and conductor lead layers is a result of the conventional manufacturing processes illustrated in FIGS. 3a-e and 4a-e.
FIGS. 3a-e illustrate the conventional manufacturing process for an overlaid MR transducer. In step 3a, the pinning, reference (either transverse bias or pinned layer), spacer, sense and overcoat films 10 are sequentially deposited on the bottom gap layer over an entire substrate wafer. In step 3b, after the depositions of step 3a, bilayer photoresists 30 and 31 are applied over the protective overcoat 32 and exposed in a photolithographic tool to mask the MR sensor in the active read region, and then developed in a solvent to form an undercut 33. The undercut 33 is provided to allow subsequent liftoff of the photoresist. In step 3c, the protective overcoat 32 in the unmasked end (or tail) regions is removed by ion milling to expose the top layer of the MR sensor. In step 3d, longitudinal bias films 12 and 13 and conductor leads 14 and 15 are deposited on top of the MR sensor. Longitudinal bias film material 36 and conductor lead material 37 is also deposited on the photoresist 31. The shadowing of the MR sensor by the photoresist 31 does not provide a vertical, aligned edge of the longitudinal bias films 12 and 13, and conductor leads 14 and 15, but, rather, provides a gradual slope of the longitudinal bias films and the conductor leads toward the central area of the sensor. Hence, the term "tail" is often used for the end region. The gradual angle of the slope makes the precise alignment of the edge difficult, thereby making a precise definition of the active sense region equally difficult. After the depositions, the photoresists are lifted off in step 3e.
FIGS. 4a-e illustrate the conventional manufacturing process for an abutted MR transducer. In step 4a, the pinning, reference (either transverse bias or pinned layer), spacer, sense and overcoat films 20 are sequentially deposited on the bottom gap layer over an entire substrate wafer. In step 4b, after the depositions of step 4a, bilayer photoresists 40 and 41 are applied over the protective overcoat 42 and exposed in a photolithographic tool to mask the MR sensor in the active read region, and then developed in a solvent to form an undercut 43. The undercut 43 is provided to allow subsequent liftoff of the photoresist. In step 4c, the reference, spacer, sense and overcoat films in the unmasked end (or tail) regions is removed by ion milling, leaving the sensor 20. In step 4d, longitudinal bias films 21 and 22 and conductor leads 23 and 24 ate deposited at the end regions, abutting the MR sensor. Longitudinal bias film material 46 and conductor lead material 47 is also deposited on the photoresist 41. The shadowing of the MR sensor by the photoresist 41 does not provide a vertical, aligned edge of the sensor 20 as the result of the ion milling. Rather, the edge forms a "tail" or slope. The shadowing by the photoresist 41 during deposition of the longitudinal bias films 21 and 22, and conductor leads 23 and 24, does not provide an aligned edge of the deposited films, but, rather, provides a gradual slope of the longitudinal bias films and the conductor leads toward the central area of the sensor. The various slopes make the precise alignment of the edges difficult, thereby making a precise definition of the active sense region equally difficult. After the depositions, the photoresists are lifted off in step 4e.
As discussed above, the trackwidth is typically defined by the edge boundary between the MR sensor and bias/leads layers. This trackwidth definition is difficult mainly due to the use of the bilayer photoresist for the ease of photoresist liftoff.
As shown in FIGS. 3d and 4d, the longitudinal bias and conductor leads films penetrate imprecisely into the undercut regions. The extent of this penetration depends on sputtering modes (RF diode, RF magnetron, DC magnetron and ion beam), shadowing effects of the bilayer photoresists, and overhangs of bias/leads films deposited on the sidewalls of the top photoresist. Thus, trackwidth is difficult to determine from the dimension of the top or bottom photoresist. In addition, the shadowing effects of the photoresists also lead to substantial reduction in the thicknesses of bias/leads films at the sensor edges, and difficulties in achieving magnetic moment balance between the read and tail (or end) regions. This moment balance is crucial in stabilizing the MR sensor without reduction in read sensitivity.